Historical hydrographic data from the continental shelf off eastern South America are used to examine the thermohaline properties of the water masses in the region between 20°S and 40°S. The continental shelf water masses are originated by dilution of open ocean waters of the western boundary currents of the South Atlantic Ocean. On the basis of temperature‐salinity relation, two distinct water masses are identified, namely, the Subantarctic Shelf Water and the Subtropical Shelf Water. Subantarctic Shelf Water originates by dilution of Subantarctic Water, primarily in the southeast Pacific, due to excess precipitation and continental runoff and enters the continental shelf near 55°S. The Subtropical Shelf Water is modified South Atlantic Central Water diluted by continental runoff from the coast of Brazil. In addition, substantial dilution of the upper shelf waters takes place at the mouth of Río de la Plata (approximately located at 36°S) and, in a lesser extent, at the Patos‐Mirim Lagoon (at 32°S). The Río de la Plata and the Patos outflows form a low‐salinity tongue that caps the shelf water leading to a salinity decrease to values <30. The low‐salinity tongue extends northward over the shelf penetrating farther north in winter than in summer. The extent of the low‐salinity water has a strong impact on the vertical stratification and acts to limit winter convection to the layer above the halocline. There is little or no indication of mixing between Subantarctic Shelf Water and Subtropical Shelf Water. An intense temperature, salinity, and nutrient front separates these water masses. The front is oriented along the north‐south direction, located on average near the 50 m isobath at 32°S and extends southward toward the shelf break near 36°S. Between 32° and 34°S the Subtropical Shelf Front follows the 100 to 200 m isobaths and separates Subantarctic Shelf Water from the oceanic South Atlantic Central Water. On the basis of the temperature and salinity distributions, beneath the low‐salinity surface layer, the Subtropical Shelf Front appears as an extension of the Brazil‐Malvinas Confluence over the continental shelf of South America. Thus the location of the Subtropical Shelf Front may be linked to the migrations of the separation point of the Brazil‐Malvinas Confluence from the continental slope.

The influence of the Plata, the second largest river in South America, extends along a coastal strip of 1300 km. Historical hydrographic and wind data and numerical simulations are combined to determine the seasonal and interannual variability of the Plata plume and its relationship to the magnitude of the river discharge and the intensity and direction of the wind stress. Our results indicate that the seasonal variability of the river plume is controlled by the alongshore component of the wind stress. During El Niño the effects of the wind and precipitation anomalies tend to compensate each other, preventing anomalous northeastward plume extensions associated to large outflow events. Numerical experiments confirm this finding and indicate that during El Niño the discharge from the Plata River spreads offshore.

The Pilot Research Moored Array in the tropical Atlantic (PIRATA) was developed as a multinational observation network to improve our knowledge and understanding of ocean–atmosphere variability in the tropical Atlantic. PIRATA was motivated by fundamental scientific issues and by societal needs for improved prediction of climate variability and its impact on the economies of West Africa, northeastern Brazil, the West Indies, and the United States. In this paper the implementation of this network is described, noteworthy accomplishments are highlighted, and the future of PIRATA in the framework of a sustainable tropical Atlantic observing system is discussed. We demonstrate that PIRATA has advanced beyond a “Pilot” program and, as such, we have redefined the PIRATA acronym to be “Prediction and Research Moored Array in the Tropical Atlantic.” Abstract The Pilot Research Moored Array in the tropical Atlantic (PIRATA) was developed as a multinational observation network to improve our knowledge and understanding of ocean–atmosphere variability in the tropical Atlantic. PIRATA was motivated by fundamental scientific issues and by societal needs for improved prediction of climate variability and its impact on the economies of West Africa, northeastern Brazil, the West Indies, and the United States. In this paper the implementation of this network is described, noteworthy accomplishments are highlighted, and the future of PIRATA in the framework of a sustainable tropical Atlantic observing system is discussed. We demonstrate that PIRATA has advanced beyond a “Pilot” program and, as such, we have redefined the PIRATA acronym to be “Prediction and Research Moored Array in the Tropical Atlantic.”